of Edinburgh, Session 1883 - 84 . 
695 
sky, and clear of all surfaces, and so that air could circulate freely 
over them. After time was allowed for them to acquire their 
respective temperatures, their readings were again taken. They 
were then slung and read, and so on, a number of readings being 
taken when slung, and when hung up. The result was that the 
thermometers read almost alike when slung, but the clean bulb 
always fell more than one degree and sometimes nearly two degrees 
when hung in the calm air. This fall was always nearly regained 
when slung ; whereas slinging did not produce any decided effect 
on the silvered bulb, the effect being so small as to be lost in the 
changes in the temperature of the air. These experiments show us 
that the silvered bulb is as suitable for night as for day observa- 
tions, and that practically it takes up the temperature of the air. 
radiation having but a small effect on it. The differential radiation 
bulbs also seem well suited for night observations; but the bulbs 
require to be coated in a different manner from those used in day 
experiments, as the quality of the heat is then different. 
Conchision. 
For travellers, these differential radiation and silver coated ther- 
mometers seem to possess special advantages, as they enable the 
observer to get nearly correct readings without screens, because he can 
be nearly indifferent to all save direct sunshine, and will be able 
with confidence to expose his thermometers anywhere, in shade, 
where there is a free circulation of air. 
One result of these experiments is to show us that, oven when 
working with our most accurate methods of observation, we have 
been regularly overestimating the temperature of our bright days 
by about one or two degrees, and in some climates this overestimate 
may be even greater. In recording future observations with the sling 
thermometer, it will be necessary to say whether the temperatures 
have been taken with clean or with silvered bulb thermometers, in 
order that the necessary corrections may be made. 
It is no very pleasant conclusion to an investigation such as this, 
to find that it ends in taking a degree or so off the average 
maximum of our summer temperature. Had it been in my power, 
it would have been far more pleasant to add to it a degree or two. 
We may, however, console ourselves with the idea that we have, 
